Probing Gelation and Rheological Behavior of a Self-Assembled

Subscriber access provided by UNIV OF NEWCASTLE

Article

Probing Gelation and Rheological Behavior of a Self-Assembled Molecular Gel Seyed Meysam Hashemnejad, and Santanu Kundu Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.7b01531 • Publication Date (Web): 17 Jul 2017 Downloaded from http://pubs.acs.org on July 23, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Langmuir is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Probing Gelation and Rheological Behavior of a Self-Assembled Molecular Gel Seyed Meysam Hashemnejad, and Santanu Kundu* Dave C. Swalm School of Chemical Engineering, Mississippi State University, MS State, MS KEYWORDS: Molecular gel; In-situ gelation; Stress relaxation; Cavitation rheology; Fracture

ABSTRACT:

Molecular gels are being investigated over the last few decades, however, mechanical behavior of these self-assembled gels is not well understood, particularly, how these materials fail at largestrain. Here, we report the gelation and rheological behavior of a molecular gel formed by selfassembly of a low molecular weight gelator (LMWG), di-Fmoc-L-lysine, in 1-propanol/water mixture. Gels were prepared by solvent-triggered technique and gelation was tracked using FTIR spectroscopy and shear rheology. FTIR spectroscopy captures the formation of hydrogen bonding between the gelator molecules and the change in IR spectra during the gelation process correlates with the gelation kinetics results captured by rheology. Self-assembly of gelator molecules leads to fiber like structure and these long fibers topologically interact to form a gel like material. Stretched-exponential function can capture the stress-relaxation data. Stress relaxation time for these gels have been found to be long owing to long fiber dimensions and the stretching exponent value of 1/3 indicates polydispersity in fiber dimensions. Cavitation

1 Environment ACS Paragon Plus

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

rheology captures fracture like behavior of these gels and critical energy release rate has been estimated to be of the order 0.1 J/m2. Our results provide new understanding of the rheological behavior of molecular gels and their structural origin.

INTRODUCTION

Among the rapidly growing body of literature on gel like materials, molecular gels formed by self-assembly of low molecular weight gelators (LMWGs) are of significant interest because of their potential applications in drug delivery1, oil recovery,2 heavy metal ion and nitrite detection,3,4 and cancer treatment.5 Unlike most common polymer gels, where monomers are covalently bonded, LMWGs self-assemble through noncovalent interactions such as H-bonding, and aromatic-aromatic interaction. Such self-assembly results in fiber like structure and physical crosslinking and/or topological entanglement of these fibers can lead to three-dimensional soft solid like material.6,7 Gelator molecules can be prudently chosen from a library of materials8,9 or can be rationally designed to render the gelation process stimuli-responsive.10 In fact, it has been shown that sol-to-gel transition can be induced by changing temperature, nature of solvent, ultrasound, and pH.11–15 With increasing number of experimental investigations on several gelators and with growing literature of computational studies, a significant understanding regarding the self-assembly process for these gels has been achieved.10,16–20 However, mechanical behavior of these self-assembled gels is not well understood, particularly at largestrain, which can lead to failure of these materials. Since, molecular gels are not formed by volume-spanning, covalently connected flexible or semiflexible polymer chains, conventional network theory based on the chain deformation mechanism cannot be directly applied. However, as the fibers are stiff in nature, as displayed by large persistence length,21 some similarities with the semiflexible, entangled biopolymer solutions can be anticipated.


Page 2 of 34

Page 3 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

To investigate the mechanical responses of molecular gels and to relate that to their structure, here, we consider a solvent-triggered gel of di-Fmoc-L-lysine in a 1-propanol/water mixture. For amphiphilic peptide gelators, most commonly, gelator molecules consist of an aromatic group (for example, Fmoc), short peptide groups, and a C-terminal group.7 Through π-π interaction between the aromatic groups and H-bonding between the peptide groups, selfassembly of gelator molecules results in fiber formation. Formation of β-sheet has been reported in some instances and helicity of the fibers has also been observed .7,22 Unlike many Fmoc containing gelator molecules, with no or smaller aromatic C-terminal group,2,14,22 di-Fmoc-Llysine is composed of identical Fmoc group at both end. Also, this molecule has multiple H-bond donors-acceptors present in the peptide groups. All these chemical moieties can cause complexity in self-assembly process and parallel and antiparallel stacking of gelators has been hypothesized.7 These kind of stackings cannot lead to fiber branching and formation of branching in molecular gels has often been considered as defect.23,24 However, density functional theory based computational study indicate that parallel and helical assembly are the preferred structural motifs for di-Fmoc-L-lysine dimers.25 The helical assembly can potentially lead to branching in fibers. IR spectroscopy has been routinely used to elucidate the self-assembly process in molecular gels, however, most of the studies have been reported on dried gel samples (xerogels) without capturing the evolution of chemical signature during the gelation process. In these investigations, for Fmoc-containing short peptide gelator molecules, secondary structures such as α-helical coil and β-sheet stacking have been captured.26–29 However, in a recent study by Ulijn and Tuttle et al. conducted on non-dried Fmoc-AA (9-fluorene-methyloxycarbonyl dialanine) gel samples, reported that the band at 1680 cm-1-1695 cm-1 in IR spectra, which was previously assigned to an


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

anti-parallel β-sheet, is due to the absorption of the carbamate group.22 In order to capture changes in IR spectra during the gelation process, potentially including all molecular interactions, i.e. solvent-solvent, solvent-gelator, and gelator-gelator, and to avoid any possible structural changes during dehydration, we collected time-lapse IR spectra during sol to gel transition. This was possible by replacing water with D2O, which allowed us to track the carbonyl stretching peak as gelation progresses. In situ dynamic rheological investigation has often been used to monitor the kinetics of gelation process.30 From these experiments, gel point can be determined rheologically by using multi-frequency oscillatory experiments using Winter-Chambon criteria.31 At gel point, the frequency dependence of both storage (Gꞌ) and loss modulus (Gꞌꞌ) become same, i.e. the phase angle become independent of frequency. At gel point, the isolated clusters of self-assembled gelators form initial continuous/percolated network. From this rheologically determined gel point, power law exponent and fractal dimension, that provide further details about the gel structure during the initial stage of gelation, can be captured. 23,32,33 As the molecular gels constitute of non-connected stiff fibers -sometime with branches - these gels can display interesting mechanical properties. For example, concentration dependence of shear-modulus (Gꞌ) depends on the type of gelator.34–38 Resemblance with densely crosslinked biopolymer networks has also been shown.35 Since topological interaction between the fibers is the origin of gel like behavior, fracture in these gels follows different mechanism than the covalently crosslinked polymer gels. To investigate further, we applied cavitation rheology technique on the gel samples. These experiments allow us to determine the critical energy release rate (Gc), which has been rarely reported in the literature for molecular gels. We found that selfassembled gels has significantly lower Gc compared with the polymer gels.39–41


Page 4 of 34

Page 5 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

In an earlier study, we have reported the solubility and gelation trend of di-Fmoc-L-lysine in different organic solvent-water mixtures.25 It has been found that di-Fmoc-lysine is readily soluble in aprotic solvents such as DMF and DMSO, whereas, in alcohols, the solubility can be promoted by increasing temperature. For example, di-Fmoc-lysine is not soluble in 1-propanol at room temperature but readily soluble at about 75°C. Higher inter-molecular H-bonding in organic solvents likely leads to poor solubility. We have noted that the conventional approach of using solubility parameters cannot capture the solubility and gelation behavior of this gelator. The simulation results show that the average binding energy between the gelator and the solvent is higher in aprotic solvents such as DMSO and DMF followed by polar solvents such as 1propanol, ethanol, and methanol, respectively. Whereas, average binding energy between the gelator and water molecules is the lowest. Average binding energy values do not capture the gelation trend correctly. To capture the gelation process of di-Fmoc-L-lysine, gelator-solvent binary mixture has been considered and the pseudo cohesive energy density (PCEDg-s) which is a square root of the gelator-solvent average binding energy divided by the corresponding molecular volume has been estimated. Further, we have determined a parameter, Λ = / , where CEDs is the cohesive energy density of the solvent. It has been shown that Λ can differentiate between the gelating and nongelating systems. Although, we have observed gelation in many water-organic solvent mixtures, here we investigate the gels formed in 1-propanol/water. These gels are very stable and display interesting mechanical responses as discussed below.


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

EXPERIMENTAL SECTION

Materials: Di-Fmoc-L-lysine (Fmoc-Lys(Fmoc)-OH) and 1-propanol (HPLC grade, 99.9% purity) were purchased from Sigma-Aldrich and were used as received. Deionized (DI) water with resistivity of 18.2 mΩ was used. Formation of gels: Solvent-triggered approach has been used to prepare the gel samples. A specific amount of gelator (di-Fmoc-L-lysine) was added in 1-propanol (typically 1300 µL) and was placed on a hot plate to increase the solution temperature to ≈ 75 °C. A transparent solution was obtained after about 10 minutes. Then, the vial was allowed to cool at room temperature (RT) for 30 minutes to obtain a solution temperature of 22 °C. During this process no precipitation of the gelator was observed. DI water (typically 3700 µL) was then added in the 1-propanol/gelator solution. The amount of gelator was varied to obtain samples with final gelator concentration of 1.7 mM, 3.4 mM, 5.1 mM, and 6.7 mM, respectively (equivalent to 1 mg/mL, 2 mg/mL, 3 mg/mL, and 4 mg/mL, respectively). Correspondingly, volume fraction of the gelator has be estimated as ϕ ≈ 0.0008, ϕ ≈ 0.0016, ϕ ≈ 0.0023, ϕ ≈ 0.0032, respectively. Here, gelator density is ≈1.278 g/cm3 (provided in MSDS). Volume fraction of solvent, ϕsol, was maintained at ≈0.26 (1300 µL 1propanol in 3700 µL DI water), considering negligible volume of the gelator (e.g. volume of gelator at the concentration of 3.4 mM is about 8.2 µL). After addition of DI water, the samples were either stored at RT (≈ 22 °C) for cavitation rheology experiments or transferred to the rheometer or IR instrument for in situ gelation study.


Page 6 of 34

Page 7 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Fourier Transform Infrared Spectroscopy (FT-IR): Thermo Scientific Nicolet 6700 FTIR instrument with attenuated total reflection (ATR) accessory was used to collect spectra at 4 cm-1 resolution by averaging over 64 scans over the range of 4000 cm-1 – 600 cm-1. A solution consisting of gelator in 1-propanol/D2O (immediately after addition of D2O) was placed on the ATR crystal equipped with a liquid sample holder and time-lapse spectra was collected. To avoid solvent evaporation, the chamber was covered with parafilm. The recorded spectrum was background subtracted. The FTIR spectra were collected at RT. Atomic force microscopy (AFM): Morphological analysis were conducted using a Bruker AFM operating in PeakForce mode with MPP-12120-10 probe. A thin layer (~1 mm) of a gel was sliced using a razor blade and was then placed on a clean microscope cover glass. The sample was then allowed to dry in a desiccator at RT for at least two weeks. AFM scans were performed with the scan rate of 0.5 Hz, at 512 × 512 pixels resolution, and also first-order flattened. Transmission electron microscopy (TEM): A JEOL LaB6 transmission electron microscope operating at 200 keV was used. Carbon coated copper grids with mesh size of 200 were placed on the freshly prepared gel for about a minute and was then removed. After carefully removing excess gel from the grid using a filter paper, samples were dried under vacuum for two days before collecting the images. Shear rheology:


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

A Discovery Hybrid Rheometer, DHR2 (TA Instruments) was used to investigate the rheological behavior. A 20 mm parallel-plate fixture with a solvent trap geometry was utilized. Samples at sol state (after water addition) were transferred to the rheometer and the evolution from sol to gel state as a function of time was captured. Small strain amplitude (γ0 = 0.1 %) was applied and multiple frequencies of 1 rad s-1, 5 rad s-1, 10 rad s-1, and 20 rad s-1 were considered. These frequencies have been selected keeping the measurement limitation of our rheometer in mind. After completion of gelation, frequency sweep experiments (γ0 = 0.1 %) followed by strain sweep (at a frequency of 1 rad s-1) experiments were conducted. Relaxation, creep, and creep recovery experiments were also conducted on the samples for which the gelation was conducted in situ on the rheometer. For all rheological experiments, gap size was set to 1.5 mm and all experiments were performed at least three times. Cavitation rheology: A custom-built cavitation rheology set-up42 was used to characterize the failure behavior of the gels. All experiments were performed after about 3 hr of water addition and on 5 mL sample volume placed in a 20 mL glass vial. Flat tip needles (Hamilton Company) with inner radius of ≈ 130 µm (gauge 26), 156 µm (gauge 24), and 205 µm (gauge 22) were used. Air was the pumping fluid and all measurements were conducted at a fixed compression rate of 2 mL/min. All experiments were performed at RT and were repeated at least three times.


Page 8 of 34

Page 9 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

RESULTS AND DISCUSSION:

Gel Formation Gels were formed using solvent-triggered approach. Figure 1 displays di-Fmoc-L-lysine chemical structure and the optical images of the sol to gel formation steps. Gelator was first dissolved in 1-propanol at ≈ 75 °C to obtain a clear liquid (Figure 1b). Gelation was initiated by adding water in gelator and 1-proponal solution. The gelator is insoluble in water, however, 1propanol and water are fully miscible. A turbid solution (Figure 1c) was obtained upon addition of water. The turbid solution gradually turned to a translucent gel in about 2 hr. Figure 1d displays a vial inversion test, qualitatively representing gel-like behavior, where the self-

Figure 1. Gel formation of di-Fmoc-L-lysine via solvent-triggered approach. a) Chemical structure of the gelator di-Fmoc-L-lysine. (b-d) Gel formation steps. (b) Dissolved gelator in 1propanol, c) turbid solution after water addition, d) inverted glass vial with gel after ~ 2 hr of addition of water. Gelator concentration is 3.4 mM, and ϕsol ≈ 0.26.


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 34

assembled structure can hold its own weight.

Gel Microstructure AFM and TEM have been used to capture the structure of the gels. Figure 2a and 2b display AFM height images at different magnifications. Height image of a glass slide without the sample is shown in Figure S1. Since 1-propanol and water have similar vapor pressure, homogeneous solvent evaporation was expected during the drying process. As a result, potential structural damage during drying of the gel was likely to be minimum. The AFM images indicate a fibrous network similar to other LMWG systems.11,34,43 As observed in Figure 2a, the microstructure appears to be heterogeneous and clusters on the order of a few microns can be observed. This observation is similar to that reported in literature for Fmoc–diphenylalanine (Fmoc-FF) system using confocal microscopy.43 Magnified view of these clusters shows the presence of fibers.

Figure 2 AFM and TEM images obtained for dried gel samples with the gelator concentration of 1.7 mM. a) and b) AFM height image of the gel samples at different magnifications, c) TEM images capturing the fiber diameter.


Page 11 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

TEM images (Figure 2c) display that the self-assembled fibers have diameter of ~10 nm, i.e., indicating that about 10 molecules, having end-to-end distance of ≈ 2.3 nm and the width of ≈ 0.85 nm, stack laterally to form these fibers.25 The previous studies on Fmoc peptide gels also reported fiber diameters of the order of 10 nanometers.34,43 However, in addition to the 10 nm fibers , AFM images (Figure 2b) also display structures with larger diameter, as high as ~ 50 nm. It is likely that these structures are bundles of multiple fibers linked by physical association. However, some bundling can also takes place during the sample drying process.

Tracking Sol to Gel Transition using FTIR Time-lapse infrared spectroscopy was utilized to probe the possible H-bonding of the amide groups between the adjacent gelator molecules during sol to gel transition. FTIR spectrum of water displays peaks

at ~ 1650 cm-1 and ~ 3400 cm-1,

44

overlapping with the relevant

characteristic peaks of the gelator molecules. Therefore, for FTIR investigations, D2O, which does not display such overlap, was used. Figure 3a displays time-lapse FTIR results during gelation, form sol state (just after D2O addition) to gel state (≈ 60 min) at a time interval of 5 min for the initial 30 minutes. Results are shown for the gelator concentration of 3.4 mM, consistent with the in situ shear rheology data presented in the next section. For reference, IR spectra of pure 1-propanol, D2O, and 1-propanol/D2O mixture (ϕsol ≈ 0.26) without the gelator are also shown in Figures 3b-c, whereas, that of pure gelator is shown in the Supporting Information (Figure S2).


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 34

Figure 3. FTIR spectroscopy data during the gelation process. a) Time-lapse IR spectra during sol to gel transition. Here, the gelator concentration is 3.4 mM and ϕsol ≈ 0.26. b) Infrared spectra of pure 1-propanol and D2O, and c) infrared spectrum of 1-propanol/D2O mixture with ϕsol ≈ 0.26. The FTIR spectra for different gelation time mostly overlap over the range of 4000 cm-1 – 600 cm-1 except in 1600 cm-1 – 1800 cm-1 and 3100 cm-1 – 3700 cm-1 regions (Figure 3 and insets). As displayed in the inset of Figure 3a, a broad peak in the 1600 cm-1 – 1800 cm-1 region can be tracked during the sol to gel transition. This peak has been assigned to the stretching of carbonyl functional group of the gelator,22 since IR spectrum of a 1-propanol/D2O solution is featureless at that region (see inset, Figure 3c). Peak position, centered around 1696 cm-1 has been found to be mostly remained unchanged during the initial 15 minutes. However, as the gelation progresses further, the peak moves to the lower wavenumber, that is to about 1691 cm-1, 1685 cm-1, and 1683 cm-1 at 20 min, 25 min, and 30 min, respectively. Beyond 30 minutes, no change in peak position has been observed, but the peak intensity increased slightly. In addition to the shift in peak position, the broader peak in the sol state became sharper at the gel state. The change in peak position, i.e., red-shift in vibration frequency, is associated with H-bonding between the


Page 13 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

carbonyl group and the amine group of adjacent gelator molecules.11,45 Also, transition from broad to a sharper peak and change in peak intensity can be linked to the transition from initially randomly distributed, non-assembled state to the assembled, and hydrogen bonded fiber like structure of the gelator molecules.22,46 It should be noted that the peak position associated with carbonyl band for the gelator in powder form (at ~ 1689 cm-1) is different from either gel state (at ~ 1683 cm-1) and sol state (at ~ 1696 cm-1). The difference in peak position has also been reported for other molecular gel systems.47 This is likely due to the different level of H-bonding because of changed sample environment. This affects the stretching vibrations in carbonyl group. In addition, as displayed in the inset of Figure 3a, the position of a broad peak centered around 3392 cm-1, remained unchanged as gelation proceeds. However, the peak intensity increases slightly as a function of gelation time. This peak can be associated with both stretching of hydroxyl group of 1-propanol and that of amine group in the gelator molecule. Due to the overlapping spectra of these functional groups, the hydrogen donating behavior of NH group in gelator molecules cannot be tracked precisely. The FTIR results clearly indicate that the Hbonding between the gelator molecules is one of the primary mechanism by which gelator molecules stack to form 1D fiber like structure. π-π stacking is another important factor but that cannot be captured using FTIR. Investigation of Sol to Gel Transition using Shear Rheology Evolution of shear moduli during the gelation process has been captured by using dynamic rheological experiments. Here, based on the Winter-Chambon criteria, a multifrequency oscillatory shear experiments at constant strain amplitude of 0.1% were performed, as the sol to gel transition took place. Figure 4a displays tan δ (tan δ = Gꞌꞌ/Gꞌ, where, δ is the phase angle, Gꞌꞌ is the loss modulus, and Gꞌ is the storage modulus) as a function of time for the frequencies of


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 34

1 rad s-1, 5 rad s-1, 10 rad s-1, and 20 rad s-1, respectively. The tan δ for all these frequencies crossover at a value of 0.3, representing a critical point, i.e, tan δc ≈ 0.3. This corresponds to the gel point, where Gꞌꞌ/Gꞌ ratio become independent of the frequency. This was determined by evaluating the frequency dependency of Gꞌ, Gꞌꞌ at different time points after the initiation of gelation and at the gel point Gꞌ ~ Gꞌꞌ ~ ωn. In other words, at that particular instance both Gꞌ and Gꞌ have a similar frequency dependency. As displayed in Figure S3, at t ≈ 10 min, such criterion is satisfied and the corresponding tan δ value is defined as tan δc. The scattered data of tan δ during the first few minutes was primarily due to the instrument measurement limitation at low torque values. Beyond the gelation point, tan δ decreases with time for all four frequencies and then leveled off, representing the completion of the gelation process. Figure 4b displays the corresponding shear stress values for different frequencies. For a given frequency, stress increased initially and then reached a plateau, representing completion of the gelation process. Beyond the gel point (corresponding to tan δc), all stress curves collapse into a single curve. The transition from sol to gel state is further evident as we monitor the values of elastic (Gꞌ) and viscous (Gꞌꞌ) moduli. As displayed in Fig. 4c, Gꞌ and Gꞌꞌ crossover occurs close to the time corresponding tan δc. Further, Gꞌ surpasses Gꞌꞌ at about 10 min (the inset is also presented in Figure S4) and then almost leveled off. Gꞌ is more than one order of magnitude higher than Gꞌꞌ after about 30 mins indicating formation of soft-solid like gel material. The results indicate that the gel point, as determined by the Winter-Chambon criteria, is slightly different that one can defined from the Gꞌ - Gꞌꞌ crossover. Note that the gel point determined by Winter-Chambon criteria indicates the formation of a percolated network structure, which can be quite different than the fully formed gel network, after about 60 minutes of gelation, as captured by AFM study.


Page 15 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 4. Evolution of rheological parameters (Gꞌ, Gꞌꞌ, shear stress, and tan δ) as a function of time during sol to gel transition for the gelator concentration of 3.4 mM. The experiments were conducted at the frequencies of 1 rad s-1, 5 rad s-1, 10 rad s-1, and 20 rad s-1 with a strain amplitude of 0.1%. a) tan δ, b) shear stress, and c) moduli as a function of time.


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 34

The critical tan δ value, i.e., tan δc, can be used to estimate the relaxation exponent, n, and fractal dimension at the gel point.31 The relaxation exponent is given by, tan δ = tan nπ/2, where 0 < n < 1. Three regions can be defined at the gel point based on the values of n; ½ < n < 1 where Gꞌ < Gꞌꞌ, n ~ ½ where Gꞌ ~ Gꞌꞌ, and n < ½ where Gꞌ > Gꞌꞌ.31 The corresponding theory was developed based on chemically crosslinked system.31,48,49 However, this formulation has also been used for physically assembled systems, such as physically associating poly(Nisopropylacrylamide) microgels (n = 0.06) and for hydrogelation of β-hairpin peptide selfassembly ( n = 0.47).50,51 In our system, we obtain, n ≈ 0.2, i.e., the sample is reasonably elastic near the gel point. Fractal dimension (df) at the gel point can be determined from the relaxation exponent. A percolation based theory for the fully screened excluded volume has been proposed to relate n and df as, =

, where d = 3 for 3D lattice.32 Here, the fractal dimension varies from

2.5 to 1.25, if the value of n changes from 0 to 1. Using the above relationship, df has been estimated to be 2.32 for our gel, which signifies compact structure near the gel point. Such a high fractal dimension at gel point might be due to presence of branched fibers and many number of short fibers at the early stage of sol to gel transition. The rheological data can be discussed in the context of FTIR data. As indicated from FTIR results, at about 20 min, we notice a shift in carbonyl peak from 1696 cm-1 to 1691 cm-1 which further shifts to 1685 cm-1 at about 25 min and then almost remain unchanged. Interestingly, as observed in Figure 4c, a clear change in slope of shear moduli has also been seen at about 20 min (also see Figure S5), matching with the FTIR data. Formation of a gel like material starting from the molecular level self-assembly of gelators to 1D fibers to 3D network is not well understood. A computational study indicates that parallel and


Page 17 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

helical assembly are the preferred, energy minimum structural motifs for di-Fmoc-L-lysine dimers.25 Additional stacking beyond dimerization lead to fiber formation and the helical assembly can potentially cause branching in these fibers. With growing number of fibers and with increasing fiber length a percolation is reached, as manifested by the gel point. Farther away from the gel point, the fibers are quite long, as observed in AFM images. Longer fibers topologically interact (entangle) to form a 3D network.6 In addition, adjacent fibers can physically interact to form bundles.6

Gel Rheological Properties and their Concentration Dependence To characterize the deformation behavior of the gel samples, frequency sweep and strain sweep experiments were performed after the gelation is complete, approximately 60 minutes after the sample loading, as tracked by the time sweep experiments. Figure 5a displays shear moduli (Gꞌ and Gꞌꞌ) as a function of frequency for the gels with different gelator concentration. The results indicate that for all samples tested here, Gꞌ is at least one order of magnitude higher than Gꞌꞌ. Also, both Gꞌ and Gꞌꞌ are almost independent of frequency, confirming soft-solid like behavior. Gꞌ has found to be increasing with increasing gelator concentration. Low-strain or linear elastic modulus, Gꞌ, can be related to the gelator concentration (C) or gelator volume fraction (φ) using a power law relationship such as Gꞌ ~ φm or [C]m, where, m is the power-law exponent. Here, we obtain Gꞌ ~ φ1.8 (Figure 5b). Note that the lowest concentration selected here is the minimum concentration below which no gel formation takes place, whereas, above the highest concentration considered here precipitation of the gelator during the gelation process has been observed. The power law exponent varies depending on the


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 34

underlying network structure and the type of gelators. Theoretical analysis for entangled semeflexible polymers/biopolymer solutions provides m = 1.4 for tightly-entangled solutions and m = 2.2 for loosely-entangled solutions.52 However, for densely crosslinked biopolymer networks m = 2.5 has been predicted,53 whereas, experimental results indicate m in the range of 2.1- 3.54–56 For colloidal gels, Gꞌ displays stronger concentration dependence, for example Gꞌ ~ [C]4.2 for a ZrO2 particles system.57 Earlier reports on molecular gels indicated, shear modulus scale with gelator concentration with m ≈ 1.8 – 4.6.34–38,58 In comparison to polymer gels, rationalization of concentration dependence of modulus for molecular gel is complex. As observed in microscopy images, fiber diameter is much smaller than its length, thus the fibers can be considered relatively stiff. Increase of gelator concentration can cause formation of new fibers resulting in increasing number of load bearing fibers. But, at the same time increase in fiber diameter, albeit small, can lead to higher bending modulus and the corresponding shear-modulus.59 The smaller exponent (m ≈ 1.8) found for our gels compared to semiflexible polymer network is likely due to the fiber branching, local heterogeneity, or presence of smaller fibers. These parameters are highly concentration dependent. i.e., structural defects increases as a function of gelator concentration.


Page 19 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 5. Concentration dependence of shear-moduli. a) Elastic modulus, Gꞌ, and loss modulus, Gꞌꞌ as a function of frequency. Gꞌ and Gꞌꞌ are represented by filled and hollow symbols, respectively, viz. 1.7 mM (■,□), 3.4 mM (●,○), 5.1 mM (▲,∆), and 6.7 mM (♦,◊). b) Elastic modulus (Gꞌ) as a function of gelator volume fraction. Dashed line represents the power law fitting. The gels were also subjected to increasing strain amplitude until the failure of gel takes place. Figure 6a displays shear-moduli as a function of strain amplitude. A strain-softening behavior for all gelator concentration has been observed, manifested by drop in Gꞌ above a certain strain amplitude. In fact, the strain-softening has been found to take place in two steps, a small decrease of Gꞌ at a strain amplitude of 1%, followed by a major drop beyond the strain amplitude of 10%. At high strain-amplitude, a cohesive fracture of the gel was observed. Figure S6 displays cohesive fracture of a gel with gelator concentration of 3.4 mM after subjected to strain sweep test. Interestingly, the gel did not melt to form liquid like material during the fracture process, as observed in many LWMG systems.60 In general, molecular gels display strain-softening behavior,21,61 where strain-softening is more pronounced at about a strain-


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 34

Figure 6. Shear moduli as a function of strain for different gelator concentration. Gꞌ and Gꞌꞌ are represented by filled and hollow symbols, respectively, viz. 1.7 mM (■,□), 3.4 mM (●,○), 5.1 mM (▲,∆), and 6.7 mM (♦,◊). a) Gels were prepared by adding water at room temperature (22 °C). b) Water was added at 37 °C. The error bars represent one standard deviation. All data were collected at a frequency of 1 rad/s. amplitude of ~ 1%. In contrary, our gel displayed mostly linear-elastic response up to a strainamplitude of 10%. The rheological properties of molecular gels considered here depend on the gel preparation conditions. Interestingly, if water is added in the gelator/1-propanal solution at 37 °C, instead of 22°C, the gels display strain stiffening behavior, i.e, the modulus increases with increasing strain-amplitude beyond a certain strain amplitude. As displayed in Figure 6b, the gel with gelator concentration of 5.1 mM displays a distinct strain-stiffening behavior. Limited stiffening behavior could be captured for the sample with the gelator concentration to 6.7 mM, as the sample underwent cohesive fracture after a small stiffening response. Note that the nature of organic solvent also plays an important role. We have shown earlier that gels prepared in DMFwater display only strain-softening behavior.25 For those samples, two-step strain-softening


Page 21 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

behavior has also been observed and the decrease in modulus was significant in the first step, above the strain amplitude of 0.3%. Strain-stiffening behavior has been observed for semi-flexible biopolymer networks such as collagen, fibrin, actin, and ionically crosslinked alginate gels at strain amplitude of 10% or lower.

30,62

Such response has been primarily linked to the stiffness of the semi-flexible chains.

Strain-stiffening behavior has also been reported in self-assembled triblock copolymer gels, at higher strain-amplitude, because of finite extensibility of flexible midblock chains.42 Another synthetic hydrogel, polyisocyanopeptide hydrogel, consisting of β-helical structure, exhibits strain-stiffening behavior allowing the applications of these gels in directing stem cell growth. 63,64

Stiff nature of these polymers has been hypothesized to be originate from the intramolecular

hydrogen bonding and packed helical structure.63,64 Strain-stiffening behavior is not very common in molecular gels. It has only been reported for a limited number of cases and the origin of such behavior is not well understood. For example, Fmoc-phenylalanine (Fmoc-F) gels display strain-stiffening behavior beyond a strain value of ~10%.14 Note that di-Fmoc-L- lysine and Fmoc-F have some similarities in molecular structure, as Fmoc-F also has an additional benzene ring. Presence of two aromatic groups likely resulted in stiffer fibers (larger persistence length) in these cases. The observed strain-stiffening behavior is most likely related to the stiffness of the fibers, similar to the semiflexible polymers. We hypothesize that to observe the strain-stiffening behavior, the fibers must have reasonably large diameter, have large persistence length, and should not slip past each other during the application of strain. All these factors in combinations dictate the mechanical responses and a change in gel preparation condition can affect one of these parameters, leading to strain softening behavior.


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 34

Note that even the strain-softening gels presented here can sustain relatively large strain value of ~10% before failure of the material.

Relaxation Behavior Applied stress or strain on a gel relaxes over time due to their viscoelastic behavior. To reveal the relaxation behavior of molecular gels considered here, stress relaxation experiments were conducted. A small step-strain of 0.1% was applied and elastic modulus, G(t), as a function of time was recorded for a duration of 1800 s. Applied strain value was lower than the typical strain value for the initiation of strain-softening or failure behavior of these gels (see Figure 6). Relaxation behavior is displayed in Figure 7 in which first 2 s of the data is not shown, as that portion of the data is noisy because of rheometer inertia effect. A decay in shear-modulus with time was observed, however, the stress did not decay completely after 30 minutes. As displayed in Figure 7, the gel relaxes about half of the initial applied stress during 1800 s.


Page 23 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 7. Stress relaxation behavior for an applied strain of 0.1%. Dashed curve is model fitting. The gelator concentration is 3.4 mM.

For self-assembled gels, the relaxation behavior as a function of time can be fitted with a

stretched exponential function as = exp ,6,65–67 where, is zero-time shear modulus, τ is the relaxation time, and β is the stretching exponent. To fit this model to our experimental data both τ and β are treated as fitting parameters in which we apply 0< G0

Probing Gelation and Rheological Behavior of a Self-Assembled

Recommend Documents